This competing continuation R01 for the Prevention Science and Methodology Group (PSMG;R01-MH40859) is to support the development of new statistical methodologies and their application in innovative settings This competing continuation R01 for the Prevention Science and Methodology Group (PSMG;R01-MH40859) is to support the development of new statistical methodologies and their application in innovative settings to the design and analysis of preventive and early intervention trials targeting mental health problems and disorders as well as drug and alcohol dependence and abuse. The analytic methods we have developed over this last five years, including growth mixture models (GMM) and general growth mixture models (GGMM), have provided researchers with a rich set of models to examine intervention impact over time. This application is to develop new analytic methods and new statistical designs to understand the developmental, contextual, and person level influences on intervention outcomes, the mechanisms by which mediation occurs, and the successful implementation and integration of effective preventive and early interventions in communities. This work directly addresses the major research questions now facing the fields of prevention and early intervention research. Our specific aims are: 1. Develop new statistical designs and analytic methods for examining intervention impact across person, environment, and developmental course. These intervention models will be used to examine multilevel mediation/moderation, impact on multiple contexts across time, and variation in impact across subgroups differentiated by their early and developing risk for psychopathology. We will test and refine these methods on existing and new prevention and early intervention trials that target conduct disorder, drug abuse, suicidal ideation and behavior, school failure, depression, and risky sexual behavior. 2. Develop new statistical designs and analytic methods for examining preventive and early intervention impact on low base rate disorders. Special statistical methods are necessary to test prevention strategies for outcomes that are rare. We will develop innovative statistical designs and analytic procedures and examine their ability to test prevention strategies for completed suicide, major depressive disorder, schizophrenia, and other major mental disorders, as well as substance dependence, and HIV/AIDS. 3. Develop new designs and analytic methods for examining impact as interventions are implemented community-wide. In the next five years, we will develop new statistical designs for stages of research that follow tests of effectiveness. These designs will address how interventions and intervention components can best be combined;whether program fidelity and program effects remain high over time, how implementation and outcome are affected as a program is scaled up in a community;and how program adaptation or tailoring to a population can best be studied. We will use existing effectiveness trial data for the prevention of conduct disorder, school failure, and drug abuse to develop and refine these models.